Revisión Crítica de Estudios de Farmacoeconomía

Presentación del tema: "Revisión Crítica de Estudios de Farmacoeconomía"— Transcripción de la presentación:

1 Revisión Crítica de Estudios de Farmacoeconomía
MSc. Arely Lemus

2 “ El saber no es suficiente. Debemos aplicarlo ”
Goethe.

3 Anatomía de una Decisión
ANATOMIA DE UNA DECISION EVIDENCIA JUICIOS RESULTADOS PREFERENCIAS DE ACCION ANALISIS CURSO CURSO Eddy dm. Anatomy of a decision. JAMA, 1990; 263:441

4 Radiografía de las decisiones con farmacoeconomía.

5 A. Búsqueda de artículos de farmacoeconomía
Búsqueda de texto libre: Economía Economics*, cost, costs, cost-effectiveness. Búsqueda en Pub-Med, Lilacs: Términos MeSH. Economics, economic value of life, costs. Cost analysis, cost-effectiveness AND (área terapéutica o medicamento que nos interesa estudiar).

6 B. Guía para estudios de evaluación económica (BMJ)
La guía tiene 10 temas: Diseño del estudio (1) Pregunta de Investigación (2) Selección de Alternativas (3) Tipo de Evaluación Recolección de datos (4) Datos de Efectividad (5) Medida de beneficio y valoración (6) Costeo Modelaje (7) Modelaje Análisis e interpretación de resultados (8) Ajustes temporales de Costos y Beneficios (9) Análisis de Incertidumbre / Análisis de Sensibilidad (10) Presentación de Resultados

7 Diseño del estudio: 1. Pregunta de Investigación
Importante en términos económicos (implicaciones de recursos) y relevante para las alternativas que enfrenta tomador de decisión ¿La promoción de la salud vale la pena? ¿vale la pena en comparación con qué? Cuando sea necesario incluir alternativas: “hacer nada” o mantener el status quo Planteamiento que considere costos y efectos ¿El medicamento X es más costoso que la terapia existente? falta información comparable de efectividad Justifique perspectiva de estudio 1. Debe ser importante en términos económicos (implicaciones en recursos) y relevante para alternativas que enfrenta tomador de decisión. Ejemplo: ¿La promoción de la salud vale la pena? NO CUMPLE CRITERIO — ¿vale la pena en comparación con qué? Incluir cuando sea necesario alternativas: “hacer nada” o mantener el status quo 2. Debe estar planteada de tal forma que considere costos y resultados. Ejemplo: ¿El medicamento X es más costos que la terapia existente? NO CUMPLE CRITERIO — falta información de efectividad comparable, puede ser q sea más costoso pero más efectivo... 3. Debe plantearse y justificarse claramente la perspectiva del estudio De la perspectiva dependen los costos y beneficios que se deben incluir El investigador debe identificar potenciales tomadores de decisión al inicio de la evaluación a quienes podría ser relevante los resultados de este estudio The research question, or hypothesis, needs to satisfy three criteria. Firstly, the question should be economically important (in terms of its resource implications) and be relevant to the choices facing the decision maker. The question "Is health promotion worthwhile?" does not meet this criterion because it fails to specify alternatives--worthwhile compared with what? Furthermore, any alternatives need to be realistic. An option of "doing nothing," or maintaining the status quo, should be included when appropriate. Secondly, the question should be phrased in a way that considers both costs and outcomes. The research question "Is drug X more costly than the existing therapy?" will provide incomplete information because the decision maker also needs to consider comparative effectiveness. Thirdly, the research question should clearly state the viewpoint of the economic evaluation, and this should be justified. Possible viewpoints include those of the provider institution, the individual clinician or professional organisation, the patient or patient group, the purchaser of health care (or third party payer), and society itself. For example, hospital and other providers may need information to help in making procurement and related technology management decisions; individual clinicians to inform patient care decisions; health insurers or purchasers to support decisions on whether to pay for a procedure or which services to develop; and patients to know the level of costs they may incur in travelling to hospital or providing informal nursing care at home. The viewpoint chosen will in turn influence both the costs included in the evaluation--for example, whether to limit these to a given department, hospital, or locality and whether patient costs are included--and the types of outcome measured--for example, disease specific outcomes or generic measures of patients' quality of life. Health economists generally advocate adopting the broader societal viewpoint when possible. This is because data can usually be disaggregated and the analysis carried out from a number of viewpoints. Also, the additional cost of adopting a broader perspective at the outset of a study is probably less than the cost of attempting to gather additional information later. Researchers should therefore identify key potential decision makers (government, purchaser, or provider) at the outset and be able to show that the research question posed will meet the needs of all key groups.

8 Diseño del estudio: 2. Selección de Alternativas
Debe plantearse claramente la hipótesis a probar, o la pregunta de investigación del estudio Debe explicarse claramente la racionalidad de la selección de las alternativas a comparar. Elección de las alternativas relevantes a comparar: La alternativa disponible más costo efectiva. La alternativa más comúnmente utilizada (en la práctica) “Hacer nada”, ¿Es un buen comparativo?. (2) SELECTION OF ALTERNATIVES * The rationale for choice of the alternative programmes or interventions for comparison should be given. * The alternative interventions should be described in sufficient detail to enable the reader to assess the relevance to his or her setting--that is, who did what, to whom, where, and how often. The choice of the alternative must be designed to help get as close a measure as possible of the opportunity cost of using the new treatment. In principle the comparator should be the most cost effective alternative intervention currently available. In practice the comparator is usually the most widely used alternative treatment. Unless current practice is "doing nothing," it is usually best not to use placebo as the comparator. Such a study could, however, if well conducted and reported, provide information for use in conjunction with studies of other treatments also compared with placebo.

9 Diseño del estudio: 3. Tipo de Evaluación
Debe plantearse el tipo de evaluación utilizado: MC, ACE, ACU, ACB Debe mostrarse una justificación clara del tipo de evaluación elegido para el análisis con relación a las preguntas de investigación planteadas. Dado que se quiere alcanzar cierto objetivo y existen X alternativas para lograrlo, cuál es la forma más eficiente de hacerlo? (3) FORM OF EVALUATION * The form(s) of evaluation used--for example, cost minimisation analysis, cost effectiveness analysis--should be stated. * A clear justification should be given for the form(s) of evaluation chosen in relation to the question(s) being addressed. There are two types of question which require the use of different forms of evaluation (see box). The first is: "Is it worth achieving this goal?" or "How much more or how much less of society's resources should be allocated to pursuing this goal?" Such questions can be answered formally only by the use of cost-benefit analysis. Looking at one intervention alone, cost-benefit analysis addresses the question of whether its benefits are greater than its costs--that is, the best alternative use of the resources. When several competing interventions are being considered the costs and benefits of each should be examined and that combination which maximises benefits chosen. The main practical problem with cost-benefit analysis is that of valuing benefits, such as the saving of life or relief of pain, in money units. However, if we are to examine whether more or less should be spent on health care, we need to find a way of comparing the costs (benefits forgone elsewhere) with the benefits of improved health and any other resulting benefits. Even when all benefits cannot be measured in terms of money, cost-benefit analysis provides a useful framework for structuring decision making problems. The second type of question is: "Given that a goal is to be achieved, what is the most efficient way of doing so?" or "What is the most efficient way of spending a given budget?" Such questions are addressed by cost effectiveness analysis, which can take one of two forms. In the first the health effects of the alternatives are known to be equal, so only the costs need to be analysed, and the least costly alternative is the most efficient. This type of analysis is often referred to as cost minimisation analysis. Secondly, alternatives may differ in both cost and effect, and a cost effectiveness ratio (cost per unit of health effect) is calculated for each. For example, given a fixed budget for dialysis, the modality (home dialysis, hospital dialysis, or continuous ambulatory peritoneal dialysis) with the lowest cost per life year saved would, if implemented, maximise the amount of life years produced by the dialysis programme. In practice, however, the selection of the most efficient mix of programmes, given a budget constraint, is more complicated: it depends on whether alternative programmes are mutually exclusive and whether the scale of programmes can be changed without changing their incremental cost effectiveness ratios. The concept "within a given budget" is also crucial. Often authors produce a ratio of extra costs per extra unit of health effect for one intervention over another and argue that a low cost effectiveness ratio, relative to other existing health care programmes, implies that a given intervention should be provided. However, judgment is still required as the resources to meet such extra costs would inevitably come from another programme, from within or outside health care. (This point is returned to in section 10.) The third category of evaluation, cost-utility analysis, lies somewhere between cost effectiveness and cost benefit analysis. It can be used to decide the best way of spending a given treatment budget or the health care budget. The basic outcome of cost-utility analysis is "healthy years." Years of life in states less than full health are converted to healthy years by the use of health state preference values, resulting in generic units of health gain, such as quality adjusted life years (QALYs) or healthy years equivalents.22 (These approaches are discussed in section 5.)

10 Recolección de Datos: 4. Datos de Efectividad
Se deben presentar detalles del diseño, parámetros y resultados del estudio Si se basa en un ensayo clínico dar detalles sobre: selección de población de estudio método de asignación de los sujetos a los tratamientos si es intención al tratamiento si es una evaluación de una cohorte tamaño de los efectos con intervalos de confianza Si se basa en un meta-análisis se deben dar detalles sobre: método de síntesis estrategia de búsqueda criterios de inclusión de los estudios en la revisión general Data collection (4) EFFECTIVENESS DATA * If the economic evaluation is based on a single effectiveness study--for example, a clinical trial--details of the design and results of that study should be given--for example, selection of study population, method of allocation of subjects, whether analysed by intention to treat or evaluable cohort, effect size with confidence intervals. * If the economic evaluation is based on an overview of a number of effectiveness studies details should be given of the method of synthesis or meta-analysis of evidence--for example, search strategy, criteria for inclusion of studies in the overview. Economic evaluation of interventions relies on the assessment of their clinical effectiveness. The data can come from a single clinical study, a systematic overview of several studies, or an ad hoc synthesis of several sources. Any limitations which weaken the assessment of effectiveness weaken any economic evaluation based on it. The gold standard for assessing the efficacy of interventions is the randomised, double blind controlled trial. This design has the highest internal validity--that is, freedom from bias. The generalisability of the study population is important in assessing the results of clinical trials and hence their suitability for economic evaluations. Factors that can limit generalisability include: differences across countries or health systems; costs and benefits resulting only from the trial protocol but which would not arise in practice; unrealistically high compliance rates; or the appropriateness of usual practice in clinical studies that compare a therapy with best usual care. Clinical data from studies employing a "pragmatic" protocol are often more generalisable and hence preferable for economic evaluation.

11 Niveles de evidencia clínica

12 Recolección de Datos: 4. Datos de Efectividad
Factores que limitan generalización de estudios: Diferencias entre países o sistemas de salud. Tasas de cumplimiento demasiado altas en términos reales. Datos clínicos de estudios que utilizan protocolos “pragmáticos” son más fáciles de generalizar y por ende preferibles en evaluación económica. Data collection (4) EFFECTIVENESS DATA * If the economic evaluation is based on a single effectiveness study--for example, a clinical trial--details of the design and results of that study should be given--for example, selection of study population, method of allocation of subjects, whether analysed by intention to treat or evaluable cohort, effect size with confidence intervals. * If the economic evaluation is based on an overview of a number of effectiveness studies details should be given of the method of synthesis or meta-analysis of evidence--for example, search strategy, criteria for inclusion of studies in the overview. Economic evaluation of interventions relies on the assessment of their clinical effectiveness. The data can come from a single clinical study, a systematic overview of several studies, or an ad hoc synthesis of several sources. Any limitations which weaken the assessment of effectiveness weaken any economic evaluation based on it. The gold standard for assessing the efficacy of interventions is the randomised, double blind controlled trial. This design has the highest internal validity--that is, freedom from bias. The generalisability of the study population is important in assessing the results of clinical trials and hence their suitability for economic evaluations. Factors that can limit generalisability include: differences across countries or health systems; costs and benefits resulting only from the trial protocol but which would not arise in practice; unrealistically high compliance rates; or the appropriateness of usual practice in clinical studies that compare a therapy with best usual care. Clinical data from studies employing a "pragmatic" protocol are often more generalisable and hence preferable for economic evaluation.

13 Recolección de Datos: 5. Medición de Beneficios y Valoración
Se debe mostrar claramente la principal medida de resultado en la EE. Ej. casos detectados, muertes evitadas, QALYS. Si se valoran los beneficios en salud: Debe mostrarse el método de valoración utilizado Ej: Valuación contingente, lotería estandarizada, etc. Debe mencionar tipo de sujetos de los que se obtuvo la valoración. Ej: pacientes, miembros del público en gral., enfermeras, etc. (5) BENEFIT MEASUREMENT AND VALUATION The primary outcome measure(s) for the economic evaluation should be clearly stated--for example, cases detected, life years, quality adjusted life years (QALYs), willingness to pay. If health benefits have been valued details should be given of the methods used--for example, time trade off, standard gamble, contingent valuation--and the subjects from whom valuations were obtained--for example, patients, members of the general public, health care professionals. **** The primary outcome measure(s) for the economic evaluation should be clearly stated: In cost effectiveness analysis benefits are usually measured in natural units. For programmes whose main effect is to extend life the usual measure is life years gained. When the main effect is on quality of life a disease specific or generic quality of life index might be used. Sometimes the benefit measure may be an intermediate marker rather than a final outcome. For example, in comparing programmes for preventing coronary heart disease reductions in blood pressure might be used. Similarly, if two antenatal screening programmes are being compared cases detected might be chosen. Such intermediate endpoints need to be justified, however, as they may be poor surrogates for final outcomes. Only a single measure can be used in the calculation of a given cost effectiveness ratio. It cannot reflect the effects of a particular intervention on both quantity and quality of life; nor can more than one aspect of quality of life be expressed. This restriction is the main limitation of cost effectiveness analysis, as other important benefits may be overlooked. Nevertheless, several cost effectiveness ratios could be calculated relating to different outcomes--but this may lead to problems of interpretation. Authors using cost effectiveness analysis should explain why they have chosen a particular outcome measure for calculation of the ratio and reassure the reader that important outcomes are not being overlooked. In cost-utility analysis the outcome is healthy years. Quality adjusted life years measure healthy years by combining data on the life years gained by programmes with a value (usually obtained from samples of patients or the population in general) reflecting the quality of those years. Two years of life in a health state judged to be halfway between death and full health would be equivalent to one year in full health. Incremental health gain is given by the difference in quality adjusted life years produced by one intervention as compared to another. Rather than obtaining valuations for each health state and then multiplying by the time spent in each, the use of healthy years equivalents requires a scenario of a specified sequence of health states and their duration. Respondents are asked how many healthy years of life this scenario is equivalent to--hence the term "healthy years equivalents." Most methods of measuring quality adjusted life years and healthy years equivalents are based on the notion of sacrifice. In economics something is not of value unless one is prepared to give up something else in order to get it. For example, using a time trade off a respondent is asked how many years of life in a health state he or she would be prepared to give up to be in full health. Using a "standard gamble" the respondent is asked to choose between a certain health state and a gamble with two possible outcomes (one worse and the other better than the health state being valued). If health benefits have been valued details should be given of the methods used- Estimates obtained by time trade off methods reflect respondents' attitudes to time as well as their attitudes to the health state being valued. Likewise, estimates obtained by standard gamble methods reflect respondents' attitudes to risk as well as their attitudes to the health state being valued. Economists are still debating which approach is most desirable.

14 Recolección de Datos: 6. Costeo
Las cantidades y precios (costos unitarios) de los recursos deben reportarse de manera separada. Ayuda al lector a evaluar la relevancia de dichos costos en su contexto Reporte de costos debe concentrarse en los rubros más importantes, cuando hay demasiadas categorías de costos. Deben plantearse los métodos de estimación de cantidades y precios (costos unitarios) Se debe poner atención en generalizaciones de estimaciones de costos, ya que los precios relativos pueden cambiar dependiendo el lugar. Debe detallarse cualquier ajuste en precios por inflación o conversión de unidades y darse la fecha de referencia de precios y tipo de cambio. (6) COSTING Quantities of resources should be reported separately from the prices (unit costs) of those resources. Methods for the estimation of both quantities and prices (unit costs) should be given. The currency and price date should be recorded and details of any adjustment for inflation, or currency conversion, given. Costing involves estimating the resources used--for example, days in hospital--and their prices (unit costs). These estimates must be reported separately to help the reader judge their relevance to his or her setting. When there are many cost items reporting should concentrate on the main costs. When economic evaluations are undertaken alongside clinical trials data on physical quantities may be gathered as part of the trial. The interpretation of resource use resulting from the trial protocol may, however, prove difficult. One view is that everything done to a patient during a clinical trial could potentially influence outcome, so the costs of all procedures should be included. On the other hand, procedures such as clinic visits solely for data collection would not take place in regular clinical care and may seem unlikely to affect outcome. Authors should consider whether the procedures followed in the trial are typical of normal clinical practice and should justify any adjustments they make to the actual observed resource use. Outside the context of a trial, estimates of resource quantities should be based on data on real patients, collected either prospectively or retrospectively from medical records. The use of physician "expert panels" to estimate resource quantities, while common, runs the risk that respondents may give inaccurate estimates or specify the resources required for ideal care, rather than that provided in practice. Prices of resources can be obtained from the finance departments of particular institutions or from national statistics, but charges (or fees) can differ from real costs. The authors of studies should comment on the extent to which the use of charges may bias their estimates. Guidelines on economic appraisal rarely discuss in detail whether the interventions being compared should be costed at marginal or average cost. Marginal costs are the additional costs of changes in the production of a service. Some authors claim the superiority of marginal costing over average costing, but this choice can be related to context and timeframe. In the short run few costs may be variable if a change in treatment is introduced, whereas over longer periods all resources, including buildings, can be switched to other uses. Thus if the study relates to a decision of a hospital manager the short run marginal costs of the various options in his or her hospital may be the relevant costs in the current budget period. If the decision relates to a matter of national policy, however, average costs may be more appropriate as these reflect the true variable costs when many services are provided in a large number of facilities across the country. Finally, the dates of both the estimates of resource quantities and prices should be recorded, along with details of any adjustments to a more recent price level. Also, attention should be paid to the generalisation of cost estimates, since relative prices and the opportunities to redeploy resources may differ from place to place.30 Currency conversions should, when possible, be based on real purchasing power, rather than financial exchange rates, which fluctuate according to money market changes.31 32

15 Nivel de evidencia para el uso de recursos en salud. (Costos)

16 7. Modelaje Debe detallarse cualquier modelaje utilizado en el estudio de EE. Ej: Árbol de decisión, modelo de regresión, modelo epidemiológico,etc. Debe justificarse la elección del tipo de modelo y los parámetros principales. Especificar qué parámetros/variables fueron modelados y cuales observados directamente Especificar supuestos del modelo. Especificar variables fueron incluidas/excluidas (7) MODELLING * Details should be given of any modelling used in the economic study--for example, decision tree model, epidemiology model, regression model. * Justification should be given of the choice of the model and the key parameters. Modelling techniques enable an evaluation to be extended beyond what has been observed in a single set of direct observations. The model will necessarily be simplified, and the extent to which the simplification is appropriate will be a matter of judgment. Modelling may involve explicit and recognised statistical or mathematical techniques. It may, however, simply bring together data from a variety of sources into a formal prespecified conceptual framework, such as a decision analysis model incorporating best available evidence from a wide variety of sources. It may be "what if" modelling, exploring what values for particular uncertain parameters would be needed for a treatment to be cost effective. Modelling may be required (a) to extrapolate the progression of clinical outcomes (such as survival) beyond that observed in a trial--for example, the progression of disease in patients with asymptomatic AIDS33; (b) to transform final outcomes from intermediate measures--for example, survival and coronary heart disease events from cholesterol concentrations34; (c) to examine the relation between inputs and outputs in production function models to estimate or apportion resource use--for example, in a cost analysis of neonatal intensive care35; (d) to use data from a variety of sources to undertake a decision analysis--for example, of screening options for prostate cancer36; (e) to use evidence from trials, or systematic reviews of trials, to reflect what might happen in a different clinical setting or population--for example, treatments for respiratory distress syndrome in preterm infants.37 The key requirements are that the modelling should be explicit and clear. The authors should explain which of the reported variables/parameters have been modelled rather than directly observed in a particular sample; what additional variables have been included or excluded; what statistical relations have been assumed or derived; and what evidence supports these assumptions or derivations. All this information may not be included in the published paper, but it should be available to the reviewer. The overall aim of published reports should be to ensure transparency so that the importance and applicability of the methods can be clearly judged (see section 9).

17 Análisis e Interpretación de resultados: 8
Análisis e Interpretación de resultados: 8. Ajustes por temporalidad de costos y beneficios Debe especificarse el horizonte temporal en el que se consideran los costos y beneficios. Horizonte temporal suficientemente amplio para captar los efectos diferenciales de las alternativas. Debe mencionarse la tasa de descuento utilizada y la justificación correspondiente. La mayoría de los analistas coinciden en que: los costos deben descontarse en cualquier estudio si el horizonte temporal es mayor a 1 año (3%-6%, 5% la más usada). Beneficios deben descontarse con la misma tasa que los costos, aunque estén expresados en valores no monetarios (ej. QALYS) Si los costos o beneficios no se descuentan, debe presentarse una explicación al respecto. Analysis and interpretation of results (8) ADJUSTMENTS FOR TIMING OF COSTS AND BENEFITS * The time horizon over which costs and benefits are considered should be given. * The discount rate(s) should be given and the choice of rate(s) justified. * If costs or benefits are not discounted an explanation should be given. The time horizon should be long enough to capture all the differential effects of the options. It should often extend to the whole life of the treated individuals and even to future generations. If the time horizon is shortened for practical reasons this decision should be justified and an estimate made of any possible bias introduced. Justifying a short time horizon on the grounds of the duration of the available empirical evidence may be fallacious.38 If the relevant horizon for the decision is long term additional assumptions may need to be made. In health care there is a still debate on discounting Most analysts agree that costs should be discounted in any study having a time horizon longer than one year. At present most recommendations seem to vary between 3 and 6%, and a common rate in the literature is 5% per year. Certainly the analyst should use the government recommended rate, probably as the baseline value, and provide a sensitivity analysis with other discount rates. It is also helpful to provide the undiscounted data to allow the reader to recalculate the results using any discount rate. Most analysts argue that health benefits should be discounted at the same rate as costs in the baseline analysis, even if they are expressed in non-monetary units, such as life years or quality adjusted life years. A zero discount rate--or one lower than that used for costs--can be introduced in the sensitivity analysis. A lower rate is advocated so as not to penalise preventive programmes and also because the results of some studies seem to suggest it.39 However, there is no a priori economic reason to favour preventive programmes and the comparisons may be between them. Imagine two programmes having the same discounted costs and the same total (undiscounted) amount of benefits, say 100 life years, but programme A obtains these benefits between years 2 and 3 and programme B between years 52 and 53. Not discounting health benefits would result in both programmes having the same cost effectiveness ratio, which seems absurd. Moreover, if the absolute benefits of programme B were 100 years and 1 day, it would be preferred--again absurdly. It is doubtful if there is enough empirical evidence on which to base a decision on the appropriate discount rate. Moreover, if the empirical argument is accepted it should also be applied to the discounting of costs. In favour of a single discount rate for costs and benefits are, firstly, consistency between cost effectiveness and cost-benefit analysis and, secondly, the idea that it is always possible to transform wealth (resources) into health at any point in time. Then, if resources are discounted, why should health not be discounted? Given the current debates about discounting, the main emphasis should be on transparency in reporting the methods used.

18 Análisis e Interpretación de Resultados: 9. Análisis de Incertidumbre
Cuando se reportan datos estocásticos se deben detallar las pruebas estadísticas utilizadas y los intervalos de confianza de las variables principales. Si no se considera la incertidumbre de manera apropiada, el lector no podrá evaluar si las conclusiones tienen sentido y si son robustas Se deben abordar en el AS las fuentes principales de incertidumbre. Hay 3 fuentes principales: Relacionada con datos observables de las variables incluidas, con la extrapolación o modelaje y con los métodos analíticos. (9) ALLOWANCE FOR UNCERTAINTY When stochastic data are reported details should be given of the statistical tests performed and the confidence intervals around the main variables. When a sensitivity analysis is performed details should be given of the approach used--for example, multivariate, univariate, threshold analysis--and justification given for the choice of variables for sensitivity analysis and the ranges over which they are varied. A recent review suggested that one in four published economic evaluations failed to consider uncertainty at all, and only one in eight handled it well. Without proper consideration of uncertainty the reader may be unable to judge whether conclusions are meaningful and robust.41 At least three broad types of uncertainty are recognised.42 Uncertainty relating to observed data inputs--When observed data have been sampled from an appropriate population standard statistical methods should be used. Typically, confidence intervals might be presented. When both costs and effects have been derived from a single set of individual patient data a stochastic approach may be used to the presentation of the confidence intervals surrounding the cost effectiveness ratio When data come from a sample attention should also be given to sample size and power. In many studies alongside clinical trials sample size may have been determined entirely by clinical endpoints. In some cases a subsample is assumed to be adequate for collecting data on resource use, but in many cases the variability in resource use data is greater than for clinical parameters, and the distribution of values is often non-normal. Attention must be paid to whether sample sizes are adequate for the economic analyses. Ideally power calculations should be presented. Uncertainty relating to extrapolation--When data have been extrapolated or modelled (see section 7) the uncertainty inherent in that process is best handled by appropriate sensitivity analysis. Uncertainty relating to analytical methods--Uncertainties may stem from the existence of alternative analytical methods. Some issues will be avoided by an explicit statement of the approach to be adopted, but others may be usefully handled by using sensitivity analysis--for example, to present results for different discount rates, or with and without indirect costs.

19 Análisis e Interpretación de Resultados: 10. Presentación de Resultados
Reportar un análisis incremental (ej. costos incrementales por año de vida ganado de las alternativas). Los resultado principales de costos y beneficios deben reportarse de manera desagregada, antes de combinarse en una razón de CE o CU. Debe presentarse la respuesta a la pregunta de investigación original; cualquier conclusión debe desprenderse claramente de los datos reportados y acompañada de las advertencias, limitaciones y reservas apropiadas. (10) PRESENTATION OF RESULTS An incremental analysis--for example, incremental cost per life year gained-- should be reported, comparing the relevant alternatives. Major outcomes--for example, impact on quality of life-- should be presented in a disaggregated as well as aggregated form. The main emphasis in the reporting of study results should be on transparency. The main components of cost and benefit--for example, direct costs, indirect costs, life years gained, improvements in quality of life--should be reported in a disaggregated form before being combined in a single index or ratio. The results of economic evaluations are usually presented as a summary index such as a cost effectiveness or cost-utility ratio. When two or more interventions are being compared in a given study, the relevant ratio is the one that relates the additional (or incremental) benefits to the additional costs. Reporting disaggregated data allows the reader to calculate other ratios that he or she sees fit.

20 C. Interpretar resultados
BENEFICIOS Análisis Económico Eliminación Automática COSTOS Análisis Económico Aceptación Automática

21 Cambio en estado de salud
Decisiones en salud A B C D E F 1 Coste neto 2 Cambio en estado de salud INTERVENCIÓN

22 Comparando costos y beneficios.
Disponibilidad social a pagar (Umbral): / AVG Puede tener potencial Costos adicionales Buen valor por la inversión “Good value for money” Medicamentos más efectivos como estatinas, ACE-inhibitors β-blockers, Terapia antiplaquetaria, antihipertensivos, etc. En pacientes en riesgo Excluido Trasplante de corazón Cáncer pulmón screening Mamografía para cáncer mama > 65 Menos efectivo clínicamente Clínicamente más efectivo This slide applies the previous Cost Effectiveness model to a number of therapies: For example, according to a cost-effectiveness analysis of data in 10 published studies, mammography screening for breast cancer every 2 years reduces death at reasonable cost for 65 to 80-year-old women who have no serious life-threatening health conditions. Other examples of “good value for money” therapies are given by most effective drugs like statins, ACE-inhibitors, ASA for primary prevention, antihypertensives, etc. in patients at risk. Patients at risk may differ according to the considered study. Hence, no comparison should be done for promotional purposes. According to a recent study published in JAMA, lung cancer screening with helical CT is unlikely to be highly cost-effective without substantial reductions in mortality, high rates of adherence, lower rates of over diagnosis, and lower costs per screening tests. The investigators concluded that “even if efficacy is eventually proven, screening must overcome multiple additional barriers to be highly cost-effective. Given the current uncertainty of benefits, the harms from invasive testing, and the high costs associated with screening, direct-to-consumer marketing of helical CT is not advisable”. References: 1. Drummond et al. Ann Int Med 1987; 107(1): 88–92 2. last accessed on 6 January 2003 3. JAMA Jan 15;289(3):313-22 Streptokinase en lugar de of tPA for IM; ASA vs no tratar en prevención secundaria Dominante (Ahorros netos) Costos menores Cuestionable 1. Drummond et al. Ann Int Med 1987; 107(1): 88–92 2. last accessed on 6 January 2003 3. JAMA Jan 15;289(3):313-22

23 Umbrales Teóricos de Costo Efectividad
(Costos +/Beneficios +)* Laupacis1 Goldman2 Mark3 Atractivo <$20 /AVG <$20 /AVG <$50 / AVG No atractivo >$100 / AVG >$100 / AVG >$100 / AVG By generating a cost per life-year gained or quality-adjusted life-year gained, one can compare the value of different interventions using a common metric. Thresholds to gauge the relative cost-effectiveness of health interventions have been proposed by both Laupacis and Goldman in two separate publications.1,2 Both of these articles suggest that an intervention with a cost per life-year gained less than $20,000 is attractive. However, treatment costs greater than $100,000 per life-year gained are not attractive as the costs appear to greatly outweigh the potential benefits. Elsewhere, it is proposed that a general consensus exists which associates an economically attractive cost-effectiveness ratio with that of costing less than $50,000 per life-year gained, while ratios above $100,000 per life-year gained are considered economically unattractive.3 Thus, compared with Laupacis and Goldman, this slightly lowers the threshold for an attractive intervention. Other commentators that have published considerations of thresholds for cost effectiveness include Califf, Lars, Mark, Cohen and Hiatky.3 References Laupacis A, Feeny D, Detsky AS, Tugwell PX. Can Med Assoc J 1993; 148: 927–929. Goldman L, Gordon DJ, Rifkind BM et al. Circulation 1992; 85: 1960–1968. Topol EJ et al (ed). Textbook of Cardiovascular Medicine. Philadelphia: Lippincott-Raven; 2002. AVG = Años de Vida Ganados * Cifras en miles de dólares 1. Laupacis A et al. Can Med Assoc J. 1993;48: 2. Goldman L et al. Circulation. 1992;85: 3. Mark. Textbook of Cardiovascular Medicine. Philadelphia, PA:Lippincott-Raven;1998

24 Utilidad de las razones de costo por AVAC en la toma de decisiones
Intervención Coste/AVAC* Terapia trombolítica con estreptokinasa vs. terapia convencional 5.768 € Fluoxetina vs. imipramina en mujeres de 30 años con un episodio previo de depresión mayor 8.414 € Warfarina vs. aspirina en el tratamiento de la fibrilación auricular (>65 años) € Captopril vs. no captopril en la supervivencia tras el infarto de miocardio (60 años) € Profilaxis general de la úlcera gástrica producida por AINES con misoprostol vs. profilaxis en mayores de 60 € Terapia trombolítica con activador del plasminógeno vs. terapia trombolítica con estreptokinasa € Terapia erradicadora después de la comprobación de H. Pylori vs. tratamiento erradicador empírico € * Transformado de US$ 1998 a Euros 2002.

25 Tabla de Posiciones Producto Nuevo?

26 D. Aspectos de equidad y juicios de valor
Los análisis de farmacoeconomía son sobre todo informativos. Es posible que nos alejemos del principio de eficiencia favoreciendo la asignación de recursos a gente más pobre y que puede beneficiarse (Alan Maynard). Si es así, al menos conoceremos el costo de oportunidad.

27 E. ¡¡ Actuar !! “Es fácil eludir tus responsabilidades, pero es imposible eludir las consecuencias de haber eludido nuestras responsabilidades” W. Churchill

28 Taller: Revisión crítica de evaluaciones económicas
Artículo: Análisis de costes del tratamiento de la diabetes mellitus de tipo 2 con insulina glargina e insulina detemir en España. Autores: Álvarez de Guisasola F, Casal Llorente C, Rubio Terrés C. Revista Española de economía de la Salud, 2007; 6(5):

29 Guía propuesta por Sacristán et al, pag. 793.

30 Discusión ¿Cual es la relevancia de este estudio aplicado al entorno en el cuál se encuentra Ud.? ¿Considera que los estudios de este tipo podrían tener una importancia para la toma de decisiones? ¿Que aspectos considera que faltaron en esta evaluación y que piensa deberían haberse incluidos? ¿Qué fortalezas encontró en el presente análisis que pudieran ser útiles para ud. a la hora de tomar decisiones?